Biological drugs, also called biologics, are therapeutics that contain one or more active substances made by or derived from a biological source [1
]. Biologics include proteins, antibodies, peptides, nucleotides and are an attractive class of therapeutic moieties for long-term medical illnesses/conditions with no other treatment available. Approximately 31% of all drugs approved by the US Food and Drug Administration (FDA) during the last 5 years were biologics [2
]. Biologics exhibit high potency coupled with target specificity, however, there are some important challenges for their broad and effective employment: (i) their transport across various epithelial layers, such as skin and mucus, is limited due to their large size, (ii) they typically suffer from poor bioavailability via the oral route and, thus, can only be administered systemically (e.g., by injection) [3
], (iii) they are susceptible to enzymatic degradation in tissues and plasma and consequently their circulation half-life is shortened.
To address these challenges, nanosized drug delivery systems have been proposed as a promising strategy for the delivery of biologics [4
]. Nanoparticles can be used as drug carriers due to their high surface-to-volume ratio (large surface area/small size) that allows for their interaction with other similarly-sized biological entities. Most common nanocarriers in the literature are based on organic materials (e.g., liposomes, dendrimers, lipid nanoparticles) [5
]. However, the mechanism of degradation of these organic nanomaterials in biological systems is not fully understood raising concerns regarding the biocompatibility of their degradation products [6
Recently, the development of novel inorganic nanocarriers has gained significant attention and is considered among the fastest growing areas of drug delivery [7
]. Phosphate-based nanocarriers have shown great potential due to their solubility in aqueous or slightly acidic conditions, their biodegradability and biocompatibility. They consist of divalent cations (usually Ca2+
]) that form ionic complexes with macromolecules, that in turn can be easily transferred across cell membranes via ion channel mediated endocytosis [13
]. Among them, calcium phosphate (CaP) nanoparticles are an important family of biomaterials in drug delivery because of their excellent biocompatibility, low toxicity, non-immunogenicity and osteoconductivity. In fact, the high biocompatibility of CaP nanoparticles renders it advantageous compared to polymeric, lipid-based, or other metal-based nanocarriers that may cause adverse effects and are not always biodegradable, and thus pose a risk of accumulation in the body [14
]. The high biocompatibility of CaP nanoparticles is further supported by the fact that CaP is approved by the US FDA and is clinically available as bone graft formulation (Gem21S®
, Osteohealth). In this formulation, CaP acts both as a carrier of a biological drug (platelet-derived growth factor) and as a bone tissue regeneration matrix. This has prompted studies with CaP as a potential carrier for several biologics such as proteins [15
], peptides [16
] and nucleotides [9
] against various diseases.
Despite the current scientific interest of nanoparticle-based therapeutics, there are still barriers for their clinical translation with the most important ones summarized below [21
(1) Complex synthesis methods of nanocarriers, that require multiple individual preparation steps, cannot be easily scaled up and reproduced leading to batch-to-batch inconsistencies.
(2) Poor loading efficiency of biological drugs in comparison to clinically relevant therapeutic doses. Many methods do not permit facile tuning of nanocarrier properties, such as specific surface area (SSA) and size, that directly affect drug loading. Typical loading values for liposomes are in the range of 1%–10%, while for polymeric nanocarriers loading values do not surpass 20% [22
]. Loading efficiency is a crucial factor determining the excipient amount in nanopharmaceutical products (i.e., nanoparticles in this case). In case of low loading, a larger amount of nanoparticles is needed in order to deliver a clinically relevant dose of the drug. However, an increase of the excipient amount might cause undesirable side effects and increase the manufacturing cost [5
(3) Stability of the biological drug due to possible loss of its biological functionality during the loading process.
Technological developments for overcoming existing barriers are important to enable the future success of nanoformulations for drug delivery.
Here, we aim to address these challenges and engineer inorganic drug delivery nanocarriers by flame spray pyrolysis (FSP). We exploit the versatility of this scalable and reproducible nanomanufacturing process for the synthesis of CaP nanocarriers with tailored properties. FSP can produce multi-component nanoparticles with high control over their specific surface area (SSA)/size, composition and crystallinity in a single step [24
]. A distinct target here is to maximize the biological drug loading on these nanocarriers. There are still large variabilities in the drug loading capacity of CaP nanocarriers described in the literature [26
], highlighting the need for an improved protocol for the loading of biological drugs on the surface of nanocarriers. We develop here such an experimental protocol for loading biological drugs (proteins and peptides) on flame-made CaP nanoparticles by physisorption, an increasingly popular route for nanoparticle biofunctionalization [27
]. We study the loading on CaP nanoparticles of two different SSA/sizes of two biomacromolecules, bovine serum albumin (BSA) and bradykinin as model protein and peptide, respectively. We therefore address challenges regarding insufficient stability, low drug loading capacity, expensive manufacturing and poor yield and unclear biocompatibility [28
In order to further implement the developed protocol in biological drugs currently in use, we utilized the LL-37 antimicrobial peptide, currently being explored in several clinical trials for its antimicrobial, wound healing and immunomodulatory properties (see ClinicalTrials.gov Identifier: NCT02225366 & NCT04098562) [29
]. LL-37 has a net positive charge at physiological conditions, is amphiphilic and can eliminate the pathogenic microbes directly via electrostatic attraction towards negatively charged bacterial membranes [30
]. Upon its immobilization on nanoparticle surfaces, LL-37 exhibits bactericidal activity against various bacteria [31
]. Braun et al. [31
] showed the effect of surface charge and SSA of mesoporous SiO2
nanocarriers on loading and release of LL-37 and they investigated their antimicrobial activity against Escherichia coli
25922. M. Vignoni et al. [35
] investigated the antimicrobial behaviour of LL-37-loaded Ag nanoparticles in skin infections and its antibiofilm formation activity. In this case, they did not observe any bactericidal effect of LL-37 for Pseudomonas aeruginosa
, Staphylococcus epidermidis
and Staphylococcus aureus
. The antibacterial effect was attributed to Ag nanoparticles, whereas LL-37 promoted proliferation of skin fibroblasts. Garcia-Orue et al. [36
] demonstrated the antimicrobial activity of LL-37 encapsulated in nanostructured lipid carriers (NLC) against E. coli
. These nanocarriers had ~73% bacterial killing and significantly improved wound healing in vivo (mice) in comparison to free LL-37. Cherredy et al. [37
] also investigated wound healing properties of LL-37 encapsulated in polymeric nanoparticles, and more precisely in poly (lactic-co-glycolic acid) (PLGA). Although PLGA-LL-37 formulation was more efficient in wound healing than free LL-37, it did not have any significant antimicrobial activity against E. coli
as even at the highest LL-37 concentration tested (5 μg/mL), the survival of bacterial cells was ~70%. To the best of our knowledge, there are no studies so far demonstrating the use of CaP nanoparticles as LL-37 carriers. Exploring the current trend of developing novel antibiotics by antimicrobial peptides, we formulate LL-37-loaded CaP nanoparticles with high loading and high reproducibility. We further study the stability of the peptide upon its loading on CaP nanoparticles against proteinase K and investigate the antimicrobial activity against E. coli
and Streptococcus pneumoniae
3. Materials and Methods
3.1. Particle Synthesis and Characterization
CaP nanoparticles doped with 5 at% europium (vs. Ca) were produced by FSP. The Eu-doping enabled the detection of the CaP nanoparticles by monitoring their luminescence. Initially, calcium and europium precursors, i.e., calcium acetate hydrate (≥ 99%, Sigma-Aldrich, Stockholm, Sweden) and europium nitrate hexahydrate (99.9%, Alfa Aesar, Kandel, Germany), respectively, were added in the solvent mixture comprised by 2-ethylhexanoic acid (99%, Sigma-Aldrich, Stockholm, Sweden) and propionic acid (≥ 99.5%, Sigma-Aldrich, Stockholm, Sweden) in 1:1 ratio and stirred under reflux for 30 min at 70 °C. Subsequently, tributyl phosphate (≥ 99%, Sigma-Aldrich, Stockholm, Sweden) was added (phosphorus precursor), after clear solution was observed, in appropriate quantity in order to obtain Ca/P molar ratio of 2.19. A high Ca/P molar ratio is chosen in order to prevent dissolution of the particles, which highly depends on Ca/P molar ratio (increased dissolution is observed with decrease of Ca/P molar ratio). The total metal concentration of the precursor solution was 0.1 M or 0.2 M. The precursor solution was delivered to the flame through a capillary tube (SGE Analytical Science, Milton Keynes, UK) using a syringe pump (New Era Pump Systems, Inc., Farmingdale, NY, USA). The solution was atomized in the FSP nozzle by oxygen gas (> 99.5%, Linde AGA Gas AB, Stockholm, Sweden) (EL-FLOW Select, Bronkhorst, Ruurlo, Netherlands) at constant pressure (1.8 bar). The synthesis of the particles was carried out at X:Y 3:8 and 8:3 where X is the ratio of the precursor feed flow rate (mL/min) and Y is the O2 dispersion gas flow rate (L/min). The spray flame was ignited by a premixed supporting flame of methane/oxygen (> 99.5%, Linde AGA Gas AB, Stockholm, Sweden) at flow rates of 1.5 L/min and 3.2 L/min, respectively. The particles were collected on a glass fiber filter (Albet LabScience, Dassel, Germany) with the aid of a Mink MM 1144 BV vacuum pump (Busch, Mölnlycke, Sweden).
Phase identification of the as-prepared powders was performed by X-Ray diffraction (XRD). XRD data were collected at ambient temperature with a MiniFlex X-ray diffractometer (Rigaku Europe, Neu-Isenburg, Germany) utilizing Cu Kα1 radiation (1.5406 Å) and operating at 40 kV and 15 mA. XRD patterns were recorded between 10 and 80° 2θ at a step size of 0.01°. Rietveld refinement was performed using the Rigaku software. The specific surface area (SSA) was determined by the nitrogen adsorption–desorption isotherms (Brunauer-Emmett-Teller, BET method,) in liquid nitrogen at 77 K using a Tristar II Plus (Micromeritics, Norcross, GA, USA) instrument after degassing for at least 3 h at 110 °C. The morphology of the nanoparticles was observed using transmission electron microscopy (TEM) in a Tecnai BioTWIN (Fei, Hillsboro, Oregon, USA) instrument operated with an acceleration voltage of 120 kV and equipped with a 2k × 2k Veleta OSiS CCD camera. For the TEM imaging, the nanoparticles were suspended in ethanol in a water-cooled cup horn system (VCX750, cup horn Part no. 630-0431, Sonics Vibracell, Newport, CT, USA) (10 min, 100% amplitude) and one drop of the suspension was deposited onto a carbon coated copper grid (400 mesh carbon film, S160-4, Agar Scientific, Essex, UK). The grid was dried at ambient temperature overnight. Size distribution and surface charge of CaP and CaP-loaded nanoparticles were evaluated by dynamic light scattering (DLS) and ζ-potential analysis, respectively, in a Malvern Panalytical Zetasizer Ultra instrument (Malvern, UK) at 100 μg/mL particle concentration. Fourier transform infrared spectroscopy (FTIR) was performed on as-synthesized CaP nanoparticles and freeze-dried CaP conjugates in an Cary 630 FTIR spectrometer (Agilent, Kista, Sweden) in a wavenumber range of 400–4000 cm−1
with a 2 cm−1
resolution. Freeze-drying was performed in a Savant SpeedVac Plus SC210A lyophilizer (Thermo Scientific, Göteborg, Sweden) equipped with a refrigerated vapor trap (RVT4101, Thermo Scientific, Göteborg, Sweden). Absorbance measurements at 225 nm were performed using an Analytik Jena Specord 210 Plus ultraviolet and visible (UV-vis) spectrophotometer (Jena, Germany). The absorbance at 225 nm (where maximum absorbance of CaP nanoparticles was observed) is monitored over time in a partially covered cuvette in order to monitor only the top suspension layer according to the method presented in Spyrogianni et al. [49
3.2. Preparation of Nanoparticles Suspension and Macromolecules Adsorption
Nanoparticle suspensions were prepared in phosphate buffer saline (PBS) at physiological pH 7.4 at an initial concentration of 1000 μg/mL. More precisely, 5 mg of the nanoparticles were dispersed in 5 mL of PBS pH 7.4 in a conical falcon polystyrene tube of 15 mL, followed by vortex-mixing for 30 s and ultrasonication for 20 min in a water-cooled cup horn system (VCX750, cup horn Part no. 630-0431, Sonics Vibracell, Newport, CT, USA). Ultrasonic energy was provided in pulses (10s on, 2s off) at 60% amplitude. As the nanoparticles are prone to sedimentation, the ultrasonication was paused every 5 min and additional vortex-mixing was implemented for 10 s in order to optimize particle dispersion.
In order to establish the experimental protocol for loading biologics on CaP nanoparticles, bovine serum albumin (BSA, ≥96%, Sigma-Aldrich, Stockholm, Sweden, MW: 66430.3 g/mol) and bradykinin acetate salt (≥98%, Sigma-Aldrich, Stockholm, Sweden, MW: 1060.2 g/mol) were selected as model protein and peptide, respectively. BSA was selected as model protein because of its stability and its wide application in evaluating sustained drug release systems [55
], whereas bradykinin is rather small and can act as model for the loading of larger ones. Further on, as a case study LL-37 antimicrobial peptide (LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES, >95%, Innovagen, Lund, Sweden, MW: 4493.3 g/mol) was loaded on CaP nanoparticles and the antibacterial efficiency of the LL-37-CaP nanoconjugates was evaluated.
Suspensions of protein and peptide-loaded CaP nanoparticles in PBS pH 7.4 (200 μL sample volume) were prepared by the addition of 100 μL of dispersed CaP nanoparticles in PBS pH 7.4 to an equal volume of protein/peptide solution of initial concentration ranging from 200 to 2000 μg/mL. The suspensions were placed on a roller mixer (Stuart SRT9D, Staffordshire, UK) for gentle mixing at 60 rpm at room temperature for 1, 2, 4, 6 and 24 h to investigate the effect of time on adsorption. The particles were separated via centrifugation at 10,000 rpm for 20 min (particle separation was confirmed by monitoring the luminescence of the CaP:Eu nanoparticles in the supernatant that was absent) and the supernatant containing the macromolecule content that has not been adsorbed was collected for quantification using a Pierce bicinchoninic acid (BCA) protein assay kit (ThermoFisher Scientific, Göteborg, Sweden) according to manufacturer’s instructions. Absorbance was measured at 562 nm using a microplate reader (SpectraMax Plus, Molecular Devices, San Jose, CA, USA) and the amount of macromolecules was calculated from a calibration curve. The amount of protein/peptide was calculated from the difference between the initial concentration and the concentration of the supernatant. Furthermore, the loaded particles were washed once with PBS and re-dispersed in PBS. The amount of macromolecule after the washing was also quantified in the supernatant after centrifugation (10,000 rpm, 20 min) and was found negligible (≤ 5%) indicating the stability of the conjugates. The suspensions of the loaded particles were stored at 4 °C and used for further experiments within a week. The loading capacity of the nanoparticles was expressed as (mass of macromolecule adsorbed onto nanoparticles in mg)/(total mass of nanoparticles in g). For each macromolecule concentration, at least three independent triplicates (from different particle suspensions and macromolecule solutions) were made and the final loading is presented as arithmetic mean ± standard deviation of the independent sets of replicate measurements.
3.3. Proteolysis and Growth Curves Analysis
3.3.1. Degradation Assay
A total of 0.5 µg LL-37, either loaded on nanoparticles or not, was subjected to Proteinase K degradation assay. Input samples were taken without Proteinase K added, as a positive control and is considered to be “100%” when calculating protein amounts. LL-37 was incubated in 20 mM Tris-HCl, pH 8.0 with a total of 20 ng Proteinase K (Qiagen, Hilden, Germany) added to the reaction. Samples were incubated at 37 °C until given timepoints where the reaction was stopped by the addition of SDS loading buffer (at a final concentration of 50 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol, 1% β-mercaptoethanol, 12.5 mM EDTA and 0.02% Bromophenol blue) and samples were then immediately boiled for 10 min at 98 °C.
Samples were loaded onto a NuPAGE 4%–12% Bis-Tris gel (Invitrogen, Carlsbad, CA, USA) and run in MES-buffer (Invitrogen) at 40 mA/gel. Since LL-37 is a small peptide, samples were run only halfway through the gel. The gels where then washed in Mill-Q water 3 times, 10 min before immersion in Imperial Protein Stain (Thermo Scientific) for 1 h. After that Mill-Q water was used to destain the gels before image acquisition on GelDoc XRS+ (Bio-Rad, Hercules, CA, USA). Bands were quantified using Image Studio Lite v. 5.2 (LI-COR Biosciences, Lincoln, NE, USA).
3.3.3. Growth Curves
The antimicrobial properties of LL-37-loaded CaPS
nanoparticles were assessed against the Gram-negative E. coli
(HVM52 strain) and Gram-positive S. pneumoniae
serotype 4 strain TIGR4 (T4, ATCC BAA-334). E. coli
was incubated under shaking at 200 rpm overnight at 37 °C in 2 mL of Luria Bertani (LB) medium. Pneumococcal strains were grown at 37 °C on blood agar plates incubated overnight with 5% CO2
. Colonies were inoculated into C+Y medium supplemented with 1% v
of a mixture of heat inactivated horse serum and glucose to optical density at 600 nm (OD600
) 0.1, grown until exponential phase (OD600
= 0.5), and again diluted to 0.05 with the aforementioned medium [52
]. Bacterial growth was monitored in honeycomb multiwell plates using a Bioscreen C MBR instrument (Growth curves OY, Turku, Finland) in which the OD600
is recorded. Each sample was analyzed in triplicate. Imaging of bacterial cultures after 2 h incubation with the CaPS
-LL-37 at 37 °C after staining with SYTO9 Green Fluorescent Nucleic Acid Stain (Invitrogen) was performed with a Delta Vision Elite microscope (Applied Precision, Uppsala, Sweden) under a 100× magnification objective. Z-stack projection images were acquired (excitation of the particles at 350 nm wavelength and emission at 570 nm wavelength and excitation of bacteria at 491 nm wavelength and emission at 516 nm wavelength). Images were acquired using SoftWoRx imaging program (Applied Precision, Uppsala, Sweden) and analyzed with ImageJ software.